The architecture of an elevator control systems normally consists of an elevator controller for each elevator to perform elevator-related signaling and motion functions and a separate group controller to perform group-related signaling and dispatching functions. Group control functions are those functions relating to the response of several elevators to hall calls. The weak point of such a system architecture is the group controller. If the group controller fails, there is no further response to group signals, such as hall calls. To guarantee further group controlling in the case of a group failure, at least a second group controller has to be provided, with additional circuitry to detect a group failure and switch through the second (redundant) group controller.
An alternative communication system is described in U.S. Pat. No. 5,202,540, "Two-way Ring Communication System for Elevator Group Control". According to this patent, each elevator controller in a multi-elevator system provides two serial asynchronous full duplex input/output channels to communicate with the next and previous elevator controllers. These two channels allow the transmission of a message in two opposite directions on a communication ring. A single interruption of the ring, via an interrupted transmission line or a disturbed elevator controller, for example, guarantees the transmission of messages to each elevator controller in at least one of the two directions. Further, using a ring architecture allows distributing the group control function across several or all elevators, so that failure of an elevator controller does not result in failure of all group control functions.
This ring communication system has advantages in robustness and system reliability but is inherently inefficient because all messages are transmitted twice and processed twice by each node, i.e., each elevator controller, on the ring. This puts a large burden in communications processing on the CPUs of the nodes. It would be desirable to find a way to use only one ring if a) a method could be found to reliably determine the health/status of each ring and to do proper switching between them and/or b) use both if necessary as originally designed but with a degradation in function to limit CPU burden.